Pergamon S0360-1315(96)00008-5 Computers Educ. Vol. 26, No. 1- 3, pp. 179-187, 1996 Copyright © 1996 Elsevier Science Ltd Printed in Great Britain. All rights reserved 0360-1315/96 $15.00 + 0.00 GENDER, COMPUTER EXPERIENCE AND COMPUTER- BASED PROBLEM SOLVING RICHARD JOINER, ~ DAVID MESSER, 2 KAREN LITTLETON 3 and PAUL LIGHT 4 ~PsychologyDepartment, Universityof Leicester, LeicesterLEI 7RH, England, 2Psychology Division, University of Hertfordshire, Hatfield, England, 3Psychology Department, The Open University, Milton Keynes, England and 4Psychology Department, Universityof Southampton, Southampton, England Abstract--This paper reports a study involving 65 children (31 boys and 34 girls) aged between 10 and I l, which further examined the effect of software type by comparing children's performance on a male stereotyped version of the software with their performance on a structurally identical, but female stereotyped version of the software. We found that girls performedworse than boys on both versions of the software and this effect persisted even when the effect of computer experiencewas removed. There was also a gender differencein the children's preference. Girls preferred the female version more than the boys and there was also a significantrelationshpbetween the girls' preferences and their performance. There was no relationship between the boys' preferencesand their performance. The implications and explanations for these findings will be discussed. Copyright © 1996 ElsevierScience Ltd INTRODUCTION There has been concern for some time that the pattern of use of computers at schools and at home is putting girls at a disadvantage relative to boys [1, 2]. Surveys have reported that girls! use computers less than boys at home [3] and at school, both at the primary [4] and the secondary level [5]. There is also recent evidence which indicates that the proportion of girls enrollinl~ for computer science degrees is falling [6, 7]. Classroom observation indicates that girls spend less time on computers and get less help from teachers, because boys tend to dominate both the teacliers' time and resources [5]. Furthermore, a number of experimental studies have reported that boys perform significantly better than girls on computer based tasks [8]. However, there are c~ther studies which have found no performance difference [9] and Underwood and Underwood suggest that performance differences can be reduced when girls are given an equal opportunity to Work on computers. Another consistent finding is that males have more positive attitudes towards computers than females [10-14]. This gender difference in attitudes can be partly attributed to the masculine image of computers and partly to the greater computer experience of males. Boys are more likely tO use a computer at home than girls, especially for playing games, and computers are generally bought for males [11, 13, 15, 16]. Research has shown computer experience (in particular having a com0uter at home) had a stronger effect on attitudes towards computers than gender [17]. An important question is whether there is anything that can be done to make computing[ and computers more attractive to girls. Most research has examined the effect of the changing social context on girls' attitudes and performance. For example, Underwood and Underwood [9] have examined whether girls benefit from working in same sex or mixed sex groups and they have found that girls perform better in same sex groups [18-20], although this has not always been found to be the case [8, 21]. However, there is other research which has investigated the effect of different types of software on girls' attitudes. For example, a study by Todman and File] [26] compared girls' attitudes towards computers after using different types of database. They found that girls were significantly more positive towards computers after using a "subjective database" than after using an "objective database". We have also conducted a series of studies on the effect of software type on girls' and boys' computer based problem solving and found equally dramatic effects. In the first study, Barbieri and Light [22] compared girls with boys solving a complex computer based problem, in the form of an adventure game called The King & Crown, and they foundthat boys performed significantly better than girls. A second study was conducted using an isomorph 179